Advanced search
2 files | 1.89 MB Add to list

Combining natural language processing and multidimensional classifiers to predict and correct CMMS metadata

Arne Deloose (UGent) , Glenn Gysels, Bernard De Baets (UGent) and Jan Verwaeren (UGent)
Author
Organization
Project
Abstract
Computerized maintenance management systems (CMMSs) contain valuable data on the maintenance opera-tions in an organization. A large part of these data consists of unstructured, written texts contained in failure notifications which are generated each time an unexpected failure occurs, enriched with structured metadata consisting of a number of labels that allow to categorize the failures, such as the type of failure, its cause or the corrective action that was taken. In this paper, we show that natural language processing techniques can be used to predict the structured metadata based on the unstructured text and even identify mislabeled notifications or ambiguous labels. Specific attention is given to the complexity that arises from the highly technical nature of the texts combined with a telegraphic writing style and heavy use of sentence fragments and abbreviations. Moreover, it is shown that exploiting dependencies between different components of the metadata, and regarding the prediction problem as a multidimensional classification problem, can improve the reliability of the predicted labels. We illustrate and test our label prediction pipeline on the CMMS data of a large pharmaceutical company.
Keywords
Natural language processing, CMMS, Maintenance, Failure logs, Machine learning, MAINTENANCE, CLASSIFICATION, STRATEGY

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.48 MB
  • KERMIT-A1-701-accepted.pdf
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 412.16 KB

Citation

Please use this url to cite or link to this publication:

MLA
Deloose, Arne, et al. “Combining Natural Language Processing and Multidimensional Classifiers to Predict and Correct CMMS Metadata.” COMPUTERS IN INDUSTRY, vol. 145, 2023, doi:10.1016/j.compind.2022.103830.
APA
Deloose, A., Gysels, G., De Baets, B., & Verwaeren, J. (2023). Combining natural language processing and multidimensional classifiers to predict and correct CMMS metadata. COMPUTERS IN INDUSTRY, 145. https://doi.org/10.1016/j.compind.2022.103830
Chicago author-date
Deloose, Arne, Glenn Gysels, Bernard De Baets, and Jan Verwaeren. 2023. “Combining Natural Language Processing and Multidimensional Classifiers to Predict and Correct CMMS Metadata.” COMPUTERS IN INDUSTRY 145. https://doi.org/10.1016/j.compind.2022.103830.
Chicago author-date (all authors)
Deloose, Arne, Glenn Gysels, Bernard De Baets, and Jan Verwaeren. 2023. “Combining Natural Language Processing and Multidimensional Classifiers to Predict and Correct CMMS Metadata.” COMPUTERS IN INDUSTRY 145. doi:10.1016/j.compind.2022.103830.
Vancouver
1.
Deloose A, Gysels G, De Baets B, Verwaeren J. Combining natural language processing and multidimensional classifiers to predict and correct CMMS metadata. COMPUTERS IN INDUSTRY. 2023;145.
IEEE
[1]
A. Deloose, G. Gysels, B. De Baets, and J. Verwaeren, “Combining natural language processing and multidimensional classifiers to predict and correct CMMS metadata,” COMPUTERS IN INDUSTRY, vol. 145, 2023.
@article{01GPB14QZFDBS0BJN3MT21V4HM,
  abstract     = {{Computerized maintenance management systems (CMMSs) contain valuable data on the maintenance opera-tions in an organization. A large part of these data consists of unstructured, written texts contained in failure notifications which are generated each time an unexpected failure occurs, enriched with structured metadata consisting of a number of labels that allow to categorize the failures, such as the type of failure, its cause or the corrective action that was taken. In this paper, we show that natural language processing techniques can be used to predict the structured metadata based on the unstructured text and even identify mislabeled notifications or ambiguous labels. Specific attention is given to the complexity that arises from the highly technical nature of the texts combined with a telegraphic writing style and heavy use of sentence fragments and abbreviations. Moreover, it is shown that exploiting dependencies between different components of the metadata, and regarding the prediction problem as a multidimensional classification problem, can improve the reliability of the predicted labels. We illustrate and test our label prediction pipeline on the CMMS data of a large pharmaceutical company.}},
  articleno    = {{103830}},
  author       = {{Deloose, Arne and Gysels, Glenn and De Baets, Bernard and Verwaeren, Jan}},
  issn         = {{0166-3615}},
  journal      = {{COMPUTERS IN INDUSTRY}},
  keywords     = {{Natural language processing,CMMS,Maintenance,Failure logs,Machine learning,MAINTENANCE,CLASSIFICATION,STRATEGY}},
  language     = {{eng}},
  pages        = {{10}},
  title        = {{Combining natural language processing and multidimensional classifiers to predict and correct CMMS metadata}},
  url          = {{http://doi.org/10.1016/j.compind.2022.103830}},
  volume       = {{145}},
  year         = {{2023}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: